Significant amounts of content (e.g., documents, records, images, audio-visual content) are stored on computing systems. It is often the case that retrieving content consumes large amounts of bandwidth and storage (e.g. distributing information across various data centers is costly). Traditional approaches for retrieving data across large distributed systems are limited to high latency in retrieving the data and high storage overhead in maintaining multiple copies of the data.